Pezhman Azizi
Aspiring Data Science and AI Professional specializing in Machine Learning, Deep Learning, and Agentic AI Systems.
About Me
Aspiring Data Science and AI professional with hands-on experience in machine learning, deep learning, Spark-based big data processing, and agentic AI system development.
I possess a strong technical foundation through an MSc in Artificial Intelligence and practical experience building scalable models, pipelines, and multi-agent workflows. I am skilled in Python, PyTorch, PySpark, Spark SQL, and GenAI tools, with a focus on ethical AI, model performance, and impactful real-world problem solving.
Education
MSc Artificial Intelligence
- Built and fine-tuned a ResNet-50 deep learning model for a 10-class road-sign dataset, achieving 97.88% test accuracy.
- Engineered a complete training pipeline on Google Colab, including augmentation, preprocessing, and hyperparameter tuning.
- Developed scalable ML workflows using PySpark and Spark MLlib.
- Designed multi-agent intelligent systems using n8n, integrating intent classification and decision-making logic.
- Performed advanced analytics using Spark SQL and DataFrames on large datasets.
Projects & Experience
Multi-Agent Customer Support Automation System
Built an end-to-end agentic workflow in n8n integrating LLM reasoning, retrieval agents, and decision logic for automated query handling. Mapped system architecture including triggers, agent responsibilities, and routing logic.
Road Sign Recognition Deep Learning Model
Developed a high-accuracy classifier using fine-tuned ResNet-50 on 15,000+ images. Implemented augmentation, normalization, and a GPU-accelerated training loop in PyTorch.
Software Development Trainee
Practiced version control, GitHub collaboration, JavaScript fundamentals, DOM manipulation, and REST API interaction. Delivered small software projects focusing on clean coding principles.
Technical Skills
Let's work together
I'm always open to discussing new projects, creative ideas, or opportunities to be part of your visions.